This document explains how to replicate all the tables and figures included in “Does School Debating Reduce Vulnerability to Misinformation? A Field Experiment in Poland” by Krzysztof Krakowski, Bernhard Clemm von Hohenberg, and Davide Morisi, published in The Journal of Politics.

The replication package includes three datasets:
	•	“misinformation_completion_by_school.dta”: This file contains aggregate data on completion rates by school and city.
	•	“misinformation_data.dta”: This file contains the main dataset used in the analyses.
	•	“misinformation_mid.dta”: This file contains data from the mid-workshop questionnaire.

The file “codebook.txt” describes all variables in these datasets.

The package also includes four do-files with Stata code:
	1.	“1_recoding.do”: Recodes variables in the main dataset “misinformation_data.dta”. This code must be run first to generate the recoded dataset used in the main analyses.
	2.	“2_analysis.do”: Replicates all figures and tables in the article and appendix (excluding Appendix F and Figure D1). This code uses the recoded dataset generated by “1_recoding.do”.
	3.	“2_analysis_appendixF.do”: Replicates all tables in Appendix F. It uses the supplementary dataset “misinformation_mid.dta” and can be executed independently of the “1_recoding.do” code.
	4.	“2_analysis_figureD1.do”: Replicates Figure D1. It uses the supplementary dataset “misinformation_completion_by_school.dta” and can be executed independently of the “1_recoding.do” code.

The analyses were conducted using Stata/SE 16.1 for Mac (Apple Silicon), Revision 11 Oct 2022.